What's New ?
Go to Recent Changes for the webpages that I've been working on lately. I'm always tinkering with the design, my notebooks have a lot of information and ideas I have not formally written about and illustrated yet.
The 2009 paper was a cleanup and expansion of the 1993 AIAA paper. Since then, I've worked on many improvements:
Ambits and Bolts . . . No Gram Left Behind . . . Reduced Rotor Velocity . . . Velocity Transformer Track . . . Capture Rail . . . Geosynchronous Construction Orbit . . . LoopPort . . . 80 km track, 30 km Station . . . Acoustic Elevators . . . Server Sky Positioning, Debris Prediction . . . Return Track . . . Lightning Protection . . . DeflectionDrag
Ambits and Bolts
Rather than "turnarounds", I shall use John Knapman's britishism "ambits". The rotor will be composed of long (10 meters?), thin (millimeters?), separable components called "bolts", using John's terminology for his own stubby and separated bolts.
The bolts will be tightly bunched throughout most of the system, but will be split into many separate small cross section streams for passage around the ambit magnets. This small cross section can scale down the ambit radius considerably.
More under the Velocity Transformer Track heading below.
No Gram Left Behind
This is a guiding principle for making cis-lunar space safe for the deployment of trillions of tonnes of assets, into the distant future. Even rocket exhaust is a form of distributed space debris; the molecules will stay in orbit until they hit something (and erode it), or light pressure sufficiently modifies their orbits to re-enter. Billions of tonnes of exhaust molecules in orbit will not cause immediate macroscopic damage and penetration, but they will damage surfaces, and the atoms they knock off those surfaces by spallation will add to the problem. Gas-phase Kessler syndrome.
So, a goal is to design missions so that make clever use of orbital dynamics to minimize the need for rocket propulsion, or fire the rockets at very high apogee, with exhaust velocities exceeding escape velocity at that altitude. That will put the molecules in solar orbit, a vastly larger volume than cis-lunar space. Alternately, use high ISP electric rockets that propel their exhaust plumes far faster than orbital velocities. VASIMR engines with argon propellant may be available someday, replacing xenon and other expensive and scarce electric propellants.
Velocity Transformer Track
Instead of the payload dragging magnetically on the rotor, a Halbach array is driven synchronously by motor windings that connect (with a pitch change) to other transformer windings that couple to the rotor in induction motor generator mode. The field velocity in the rotor lags its synchronous speed by tens of meters per second; while this results in very high thrust, and very brief high current pulses in the rotor windings, the slip is very small and the motor efficency very high. This velocity transformation greatly reduces rotor heating, and increases launch efficiency above 99%. However, it does create stress on the track, requiring longitudinal strain relief running the length of the track.
VelocityTransformer page, with a cross section of the segmented bolt rotor.
Capture Rail
Some (needs reference!) have written about a tether hanging from a heavy orbiting station. An ascending vehicle matches velocity with the bottom of this rail, then climbs up, perhaps powered with acoustic energy on the tether. This will rob the station of momentum and angular velocity, but that can be added back with slow, high efficiency electric engines (VASIMR engines with cheap argon propellant, perhaps).
A CaptureRail climbing capture tether arrives with a higher velocity, and decellerates all the way up a relatively stiff rail. That can require no climb energy, and slightly less angular velocity correction at the station.
However, the LoopPort described below is probably a better idea.
Geosynchronous Construction Orbit
Before a capture system is built in a synchronous location in GEO, or for the assembly of interplanetary missions requiring a lower perigee, we can launch to multiple highly elliptical 86164 second period Construction Orbits with a perigee of 8378 km (2000 km altitude) and a perigee of 75950 km. These orbits are synchronized with the daily availability of the launch loop (within a 15 minute window) for the delivery of multiple payloads.
Interplanetary missions should start at a low perigee for maximum escape velocity. A low perigee also permits a "12 hour return" for crew rotation and medical emergencies given a relatively small 114 m/s delta V at apogee. Two more delta V's will put the vehicle into a circular geostationary orbit. However, for many missions, including crew missions, construction orbit is the final destination, or the launch point (at perigee) for interplanetary missions.
LoopPort: Construction Orbit Spaceport
A short spaceport resembling the 1979 Arnold/Kingsbury horizontal capture spaceport. The LoopPort is in a geosynchronous (but not geostationary!) high eccentricity "24 stellar hour" construction orbit, and captures incoming vehicles launched from the launch loop at perigee, with a relative delta V of 114 m/s. Like the LEO spaceport, the vehicles will be captured "retrograde", shedding energy and accelerating to the slightly higher LoopPort apogee velocity. They will decelerate relative to LoopPort; this can generate some energy, but it is probably cheaper to waste it, or use it to power passive stabilization coils that keep the vehicle centered (mad handwaving here).
A second stream of much smaller vehicles, carrying "dumb" payload that can tolerate high acceleration, is also launched from the loop into into a much higher orbit 12 hours later, putting it in an orbital plane tilted 16 degrees from the main construction orbit. At apogee, rockets boost velocity, angular momentum, and perigee to the apogee radius of the first stream. At plane crossing, another rocket firing matches the orbital plane to that of the first stream, and the vehicles arrive at LoopPort altitude to be captured prograde and restore angular momentum.
The velocity ratios and mass ratios are about 20 to 1.
80 km main acceleration track, 30 km altitude West Station
With a pointy nosecone, the hypersonic drag for an acceleration path at 80 km is acceptable, the lower altitude reduces stabilization cable weight, and the 33 times higher atmospheric density should reduce space debris flux by a corresponding factor.
Vehicle drag power and heating is proportional to velocity cubed. Vehicles will increase acceleration slowly from West Station, and increase speed as they gain altitude and encounter less drag. A lower altitude west station allows it to be heavier, makes maintenance and staff "commuting" easier, and reduces the time and expense of the west station elevators.
Acoustic Elevators from the Surface to West Station
I presented this at the 2017 ISEC Space Elevator Conference (ADD LINK to paper). If a very strong tether material ever becomes available, it will be stiff enough to make a superb "tin can telephone" high power vibration transport to an ascending climber.
- Sadly, carbon nanotubes are superlubricious (ADD LINK to Huang paper) and will never be both strong and crosslinked enough to make a Space elevator to GEO and behond. Graphite is a lubricant. However ...
A 30 km high elevator to West Station can be made from Kevlar, or from stronger (but more expensive) materials like Torayca T1100G ( ADD LINK to datasheet ). A Kevlar acoustic elevator must be much stouter, but can lift a vehicle in a few minutes. A launch loop will need dozens of these elevators.
This replaces recent "loop and pulley" elevator approaches.
Server Sky Positioning, Control Information, and Precision Debris Prediction
Return Track Bolt Acceleration and Deceleration
The return track isn't doing much, besides providing a platform for radars and measurement gear, and helping to support surface cables. It will carry a small track that will accelerate replacement bolts retrograde to very nearly rotor velocity, sending them down through a "13th channel", past the vertical deflectors, and into a horizontal switchyard where they can be swapped with defective bolts in the main stream. A sparse stream of replacement bolts may be maintained in a racetrack that loops through the west station bolt switchyard, so that bolts can be swapped as soon as a defect is detected.
Incline Lightning Protection
The inclines will be hit by lighting strokes. They should be sheathed in at least 2 kg/m of aluminum to reduce return resistance. The bolts in the rotor stream should be electrically isolated from each other (method TBD) so that they can capacitively couple to the outer tube and develop similar voltages. Hands again waving madly.
Deflection Drag
A Launch Loop can be arranged many ways; for this note, assume a 20 degree incline with a short subsurface section (reducing the altitude of the upward deflectors), and a raised return track. A 3000 km long loop will include 3000/6378 or 0.47 radians of track deflection in each direction. Each of four "up-down" deflectors will add around 60 degrees of deflection to the essential 360 degrees of loop "racetrack".
The total deflection angle is 4*60+360 = 600 degrees added to 2*0.47 radians, about 12 radians of total deflection θ. Moving at v = 12.5 km/s, the rotor completes one cycle in t = ( 6000 km / 12.5 km/s ) = 480 seconds, 8 minutes.
The total system deflection force is ρ v² θ, where ρ is the rotor mass density, 4.32 kg/m for the 2018 design. The deflection forces may create some drag on the rotor, characterized by a LD lift-to-drag ratio, so the system drag force is ρ v² θ / LD. The drag loss power P is the force times the rotor velocity v, or P = ρ v³ θ / LD. This does not include residual gas drag between the rotor and the track/containment. So, the deflection power P = 100 Terawatts / LD. Obviously, LD must be VERY high, or the deflection losses will be totally unacceptable.
The canonical launch loop system assumes DC magnetic deflection with very small/fast stabilization tweaks to the currents generating the magnetic field; delayed tweaks mean exponential growth of perturbations. These tweaks are NOT computed merely from the local differential spacing, but are influenced by the absolute position of the entire rotor and track. The rotor must "enter the bullseye" of each massive deflector nearly perfectly after making a 2000 km transit.
This may seem absolutely impossible, but LIGO uses lasers to measure differential distances to 23 decimal places. A "mere" 7 decimal place accuracy for a launch loop (200 micrometers at 2000 km distance) is challenging but not intrinsically impossible, especially with orbiting space assets to assist with distance measurement of the high-altitude sections, and a "laser truss" between launch and return tracks. This cannot be done without sophisticated electronics and system-global computation and optimization.
It certainly cannot be done with lossy and localized spacing control, which is characteristic of Inductrack as I understand it. A lift-to-drag ratio of 500 might be excellent for a sub-mach ground transportation system, but it is far too high (loss = 20 GW!) for deflecting a 26,000 tonne rotor moving 100 times as fast and "circling twice" in 8 minutes.
No electronics and computation, no launch loop. Fortunately, we've had electronics and computation capable of the task since 1980, which has improved 4 orders of magnitude in 2018. My Fluke benchtop voltmeter (built in 1980) is 7 decimal places. Loop stabilization electronics will be very expensive for prototypes, but integrated circuit mass production and automated assembly can make very large systems relatively cheap, especially after years of development for highly profitable Power Storage Loops.